TC bogus bug: old version output with no soil data cannot replace met_em.d*

[lzhenn]

MODEL VERSION: v4.1.2

TYPE: bug (maybe)

KEYWORDS: output, soil layer, ERA5

SOURCE FILE: /main/tc_em.F

DESCRIPTION OF PROBLEMS:
The default of the output of tc.exe does not match the met_em.d* file structure in v4.1.2, I tried to use force_use_old_data = .TRUE., real.exe can run but wrf.exe stopped after a while. With message:

Flerchinger USEd in NEW version. Iterations= 10

I then found no soil moisture/temp data in tc.exe generated aux file. After checking the code in tc_em.F, I found that tc.exe use SM000010 and grid%flag_sm000010 etc. to check soil data, but met_em data from metgrid.exe when using ERA5 got SM000007 etc, which means the soil layers are not consistent.
I try to modify the code in tc_em.F accordingly:

flag_name(1:8) = 'SM000007'
CALL wrf_get_dom_ti_integer ( idsi, 'FLAG_' // flag_name, itmp, 1, icnt, ierr )
IF ( ierr .EQ. 0 ) THEN
    grid%flag_sm000007 = 1
end if

I found flag_sm000007 etc. were defined but not used in share/module_optional_input.F. Even use flag_sm000010, the new compiled tc.exe still cannot produce correct soil mois/temp output.

[anbaichao]

The same problem